Models of functional maturation in human cortical microcircuits

The human brain undergoes substantial developmental changes from childhood to adulthood, but the underlying biophysical mechanisms and their implications for cognitive maturation remain unclear. This computational neuroscience project addresses this gap using a unique multi-modal dataset of >1,100 patch-clamp recordings from human supragranular pyramidal neurons spanning a large age range (0.9-55 years). The main goal is to understand how developmental shifts in neuronal biophysical properties influence information processing and learning in human cortex. The approach uses a new Python toolbox that allows gradient-based optimization of biophysically detailed neural models. The project has three integrated phases: first, developing age-specific biophysical models of individual cortical neurons; second, constructing circuit models by connecting these detailed neurons through biologically realistic synapse models fitted to multi-patch synaptic recordings from human cortex; and third, integrating these components into larger networks trained on cognitive tasks to compare learning dynamics between children and adults. This represents the first application of gradient-based optimization to human cortical data across development. The project has the potential of providing insights into critical periods when neurodevelopmental disorders emerge, effectively bridging computational modeling with human developmental neurobiology.